By ToAI Team
Over the past five years, artificial intelligence has achieved remarkable strides in nearly every of its traditional subfields. This includes, but is not limited to, vision, speech recognition and generation, picture and video generation, natural language processing (understanding and generation), multi-agent systems, planning, decision-making, and many more.
The article also discusses a few other AI developments, including Collibra’s new tool for AI governance, the US government’s focus on AI legislation, and concerns about proposed AI guidelines in Canada.
Collibra Unveils AI Governance Tool
At the Data Citizens conference in Orlando, Florida, Collibra introduced its product, Collibra AI Governance. With the goal of empowering users to oversee the AI models they implement, this groundbreaking product alleviates concerns regarding the potential risks associated with AI adoption while maximizing returns.
AI Governance aims to provide a solution by maximizing transparency regarding AI models’ inner workings and streamlining the governance processes that surround them. AI Governance, a component of the Collibra Data Intelligence Platform, allows users to oversee AI models and data administration controls.
The Data Intelligence suite’s data security features, in conjunction with the Collibra data catalog, ensure that all training data used to train AI models is secure and private. Implementing this cohesive integration enables the control of data access pertaining to confidential personal data.
In addition, AI Governance provides a forum for data science and AI teams to collaborate on the development of use cases, document specifics, and generate ideas for AI applications. Furthermore, it underscores potential hazards linked to these applications, including but not limited to ethical considerations, intellectual property concerns, privacy infringement, and regulatory compliance obligations.
Collibra AI, which was recently released alongside the Data Intelligence Platform, incorporates AI functionalities designed to simplify data quality and governance duties. Collibra AI now allows users to create SQL-based data quality criteria using large language models (LLMs). Additionally, it will aid in data curation by automating the description of data assets using the language generation capabilities of LLM.
US Government Focuses on AI Regulation
In an effort to mitigate the security risks that arise from the advanced technological capabilities of artificial intelligence, both the federal government and industry in the United States have been active in developing solutions. There is an increasing prevalence of regulatory discourse pertaining to artificial general intelligence (AGI).
The United States Department of State initiated the ‘Defense in Depth’ strategy to strengthen the safety and security measures surrounding artificial intelligence. House Chairman Comer and Ranking Member Raskin unveiled the plan on March 5, 2024, outlining five strategies to ensure the consistent expansion of AI technology.
Level 1 focuses on establishing protocols to sustain the advancement of artificial intelligence (AI), whereas Level 2 delineates strategies to improve preparedness for promptly addressing AI development and implementation incidents. Level 3 proposes increasing the nation’s financial commitment to AI safety research and standardization. In contrast, Level 4 underscores the incorporation of responsible AI development and implementation protocols. The United States Federal government establishes a regulatory agency and legal accountability framework for artificial intelligence, laying the groundwork for establishing AI transparency.
Concerns about Proposed AI Legislation in Canada
Introduced in 2019, the government’s Directive on Automated Decision-Making aimed to enhance transparency in using AI and algorithmic systems by conducting impact evaluations. As of now, only 18 assessments have been released to the public, notwithstanding this endeavor. Interestingly, Canadians do not trust AI technology, despite their government being one of the first to adopt an AI strategy.
The government primarily focuses on incorporating AI into industrial policies, with consultations typically perceived as symbolic rather than meaningful. Despite ongoing concerns about facial recognition systems, this method has allowed mainstream acceptance of AI technologies to continue. Furthermore, the government’s procurement methods have tended to ignore the consequences of AI implementation, seeing it primarily as a solution for enhancing Canada’s service-based public sector. It is critical to note that, while AI provides benefits, its societal ramifications are also substantial.
The Canadian Union for Public Employees, the Professional Institute for Public Employees, and the Canadian Labour Congress have all urged that AIDA be applied to government departments, agencies, and crown corporations. It is believed that AIDA was hastily implemented without thorough input from the public, which does not meet the needs of Canadians and lags behind the regulatory approaches of other countries in terms of AI. Therefore, the Canadian Union for Public Employees and the Professional Institute for Public Employees have proposed separating AIDA from Bill C-27 and sending it back for revisions.
Conclusion: Navigating the Evolving Landscape of AI
Increasing demands are being made for regulations in the field of AI. It is imperative to prioritize the supervision and regulation of AI technologies, emphasizing the significance of collaboration among technology firms, governmental entities, and the general public in order to guarantee the global progression of AI. Opposing AI should not be the objective; rather, establishing ethical regulations for it should be the aim.
What is your reaction?
